feat(desktop): import cookies from Safari by juliusmarminge · Pull Request #7262 · pingdotgg/t3code

juliusmarminge · 2026-08-16T21:31:54Z

Adds Safari as a browser import source on macOS.

Safari keeps its cookies unencrypted in Cookies.binarycookies; the protection is Full Disk Access (TCC), not a key. The reader parses the binary format with every declared structure bounds-checked and refuses a jar the page table does not describe rather than importing part of it. Timestamps are rebased from the 2001 epoch, and host-only vs domain cookies go through the same cookieScope rule as the Chromium and Firefox readers.

TCC denies the read with EPERM while still allowing stat, so listing probes the jar by opening it and reports needsFullDiskAccess up front. The wizard gains a Full Disk Access step: Open System Settings deep-links to the Privacy & Security → Full Disk Access pane through a dedicated openSystemSettings IPC method keyed by pane id (the generic openExternal allowlist is unchanged), and after granting the wizard rechecks or retries the selected import. Only Safari's primary profile jar is read; additional Safari profiles are out of scope.

Validation: binary parser fixtures and malformed-record boundaries, an EPERM-vs-EACCES distinction test, the fixed System Settings destination, and wizard state transitions. The parser was also run against reference binarycookies fixtures and structurally checked against a live Safari 27 jar.

Original implementation: Claude Code. Review fixes: GPT-5.6 Sol agents via Codex, and Claude Fable 5 via Claude Code.

Note

Add Safari cookie import with Full Disk Access permission flow

Adds Safari as a Darwin-only browser import source that reads Safari's binary cook cookie jar instead of SQLite databases used by Chromium and Firefox
Implements parseBinaryCookies in SafariCookies.ts to decode the big-endian page table, little-endian records, Safari epoch timestamps, URL/domain/path scope, and secure/HTTP-only flags; structurally invalid jars fail with SafariCookieReadError
Discovers named Safari profiles from SafariTabs metadata with fallback to scanning UUID-named WebKit data stores when metadata is unreadable
Adds a needsFullDiskAccess unavailability reason and a wizard step that lets users open the macOS Full Disk Access System Settings pane, recheck permission, and resume the import
Safari source listing returns needsFullDiskAccess when the cookie jar exists but opening it yields EPERM; EACCES and other errors remain generic read failures

^{Macroscope summarized c6cb164.}

Note

Medium Risk
Touches macOS TCC-sensitive cookie reads and new IPC that opens fixed System Settings URLs; parser mistakes could drop or mis-import cookies, but URLs are allowlisted and malformed jars fail closed.

Overview
Adds Safari as a macOS browser-import source, reading Cookies.binarycookies with a bounds-checked parser that maps records into preview cookies (epoch rebase, host-only/cookieScope, secure/httpOnly flags) and surfaces TCC denials as needsFullDiskAccess (EPERM) rather than generic read failures.

Availability and import probe Safari by opening the jar (since stat can succeed without Full Disk Access), wire the Safari engine through BrowserImport, and extend source discovery to list the default jar plus named profiles from SafariTabs.db / WebKit data stores, resolving Cookies.binarycookies per selected profile. Safari is treated as safe to import while running (no lock probe).

Desktop shell / UX: new openSystemSettings("full-disk-access") IPC (pane id, not renderer URLs) deep-links to Privacy & Security → Full Disk Access; the import wizard gains a Full Disk Access step with recheck/retry, and Integrations calls the local API to open settings. Contracts add the safari source id and needsFullDiskAccess copy; user docs note the FDA flow.

^{Reviewed by Cursor Bugbot for commit c6cb164. Bugbot is set up for automated code reviews on this repo. Configure here.}

coderabbitai · 2026-08-16T21:32:09Z

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

🔍 Trigger review

_{Comment @coderabbitai help to get the list of available commands.}

github-actions · 2026-08-16T21:35:42Z

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.5 KiB	13.8 KiB	+225 B (+1.6%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.0 KiB	7.0 KiB	−1 B (−0.0%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.5 KiB	6.7 KiB	+226 B (+3.4%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	57.0 KiB	58.5 KiB	+1.5 KiB (+2.6%)	66.4 KiB	✅
Codex	Live turn messages	8	10	+2 (+25.0%)	21	✅
Claude	Total thread wire	13.7 KiB	13.6 KiB	−172 B (−1.2%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.0 KiB	7.0 KiB	+4 B (+0.1%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.7 KiB	6.5 KiB	−176 B (−2.6%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	59.3 KiB	57.8 KiB	−1.5 KiB (−2.5%)	66.4 KiB	✅
Claude	Live turn messages	10	9	−1 (−10.0%)	21	✅

Baseline: 5e1c23d · PR result: c6cb164 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

Codex decoded thread snapshot: 113.8 KiB
Claude decoded thread snapshot: 114.5 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

macroscopeapp · 2026-08-16T22:27:59Z

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — This PR adds a substantial macOS Safari cookie-import workflow that reads protected browser data, transfers authentication cookies, and introduces new parser, permission, UI, and IPC behavior. It also adds a static-analysis suppression directive, so the scope and review-sensitive changes require human review.

Not approved because:

Per-PR cost limit exceeded (workspace setting). Approvability relies on correctness review in order to determine eligibility

^{Review your spending limits in Billing settings, or comment @macroscope-app review this PR to bypass the limit and review now. You can add or adjust custom eligibility rules. Learn more.}

macroscopeapp

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

macroscopeapp

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

cursor

Stale Bugbot comment from a previous run.

macroscopeapp · 2026-09-03T23:31:32Z

Macroscope skipped reviewing this pull request. Per-PR cost limit exceeded (workspace setting).

Reviews on this PR have cost $48.44 so far. This review would add an estimated $1.85, bringing the total to $50.29 — above your per-PR limit of $50.00.

Tip

To get this pull request reviewed, you can:

Comment @macroscope-app on this PR to request a manual review (monthly spend limits still apply).
Exclude large or generated files from review by adding a pattern to your .macroscope/ignore.md — note that creating this file replaces Macroscope's built-in default ignores rather than extending them.
Raise your cost limit in your workspace billing settings.

^{Turn off this reminder going forward}

Safari does not encrypt its cookies. It stores them in a proprietary `Cookies.binarycookies` file whose protection is TCC rather than cryptography: the file sits inside the app container, which only apps with Full Disk Access may read. So the gate is a permission the user grants in System Settings, and a denial is reported as exactly that rather than as a generic read failure. Two details the format forces: Timestamps count seconds from 2001-01-01, not the UNIX epoch, so every expiry needs rebasing or cookies import as long expired. The format predates SameSite and carries no equivalent field. Imported cookies are marked Lax, the modern browser default — claiming "none" would widen the scope of every cookie Safari ever set. Safari keeps one jar for the whole app rather than per-profile, so it exposes a single implicit profile, and it has no observable lock file since the jar is written atomically. The parser is covered by tests that build the binary format byte for byte, including a multi-page file — Safari pages its jar, and a single-page reader would silently return only the first slice. That coverage matters because the real file cannot be read on this machine without the very permission the feature asks for; the TCC path itself was verified against the live file, which denies with EPERM and reports `needsFullDiskAccess`. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

… of it `Buffer.subarray` clamps rather than throwing, so every declared structure in the binary format was taken on trust. An overlong page swallowed the following page's bytes and pushed the cursor past the end, dropping every cookie after the boundary from an import that still reported success. A record whose declared size overran its page left its string offsets free to read the next record's bytes as this cookie's value. Pages, records, and string offsets are now bounds-checked against what the file actually contains, and a mismatch fails the read. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Matches the Chromium and Firefox readers: the failure carries which jar it was for, so a Full Disk Access refusal is traceable rather than anonymous. Optional because the parser raises before a path is in hand. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

A Safari import with no Full Disk Access failed with the generic "cookie database could not be read" instead of telling the user to grant access — and no prompt appears, because macOS never prompts for Full Disk Access; the app is added by hand. The denial arrives as EPERM, which Effect tags `Unknown`, not `PermissionDenied` (that is EACCES), so checking the tag alone never matched. The underlying errno is checked too. Verified against the real jar: the reason is now `needsFullDiskAccess`, which the renderer maps to the System Settings instruction. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Safari's cookies sit behind Full Disk Access, which no one has granted before their first import — so it is a step in the flow, not a failure. When an import comes back needing it, the wizard shows a screen that says what it's for, links to the right System Settings pane, and — from an "I've turned it on" button — runs the import itself, so the user never restarts from the menu. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The step's "Open System Settings" button did nothing: it routed the `x-apple.systempreferences:` deep link through `openExternal`, whose allowlist only passes http(s) and remote-editor schemes, so the link was silently dropped. Adds a dedicated `openSystemSettings(pane)` path instead. The renderer sends a known pane identifier, not a URL, and the main process maps it to an app-fixed deep link and opens it directly — so the general link allowlist stays locked down while this one trusted destination gets through. Uses the post-Ventura Full Disk Access anchor. Also reworks the step's copy to say why the permission is needed and that it can be revoked once the import is done. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Two review findings. Safari's parser interpolated a bare IPv6 host into the cookie URL (`http://::1/`), which Electron rejects; hosts with a colon are now bracketed, matching the Chromium reader's cookieScope. And the cookie-database candidate probes accepted any stat-able entry, so a directory squatting on `Network/Cookies` would be picked, fail the SQLite open, and shadow a valid legacy `Cookies` file behind it; every candidate probe now requires a regular file, in the listing, the counts, and the importer alike. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

…d route only TCC to Full Disk Access Two review findings on the Safari reader. The parser validated each declared page but never checked that the declared pages account for the file, so a jar with extra undeclared pages imported "successfully" with those cookies silently missing; it now requires the remainder to be exactly what Safari writes after the pages — nothing, the 8-byte checksum, or checksum plus a length-prefixed property list — and refuses anything else. And `isPermissionDenied` sent ordinary EACCES failures (mode bits, ACLs) to the Full Disk Access grant, which cannot fix them; only TCC's EPERM routes there now, and the TCC test injects a real EPERM instead of relying on chmod's EACCES. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

…ch other A record offset was only checked against the page's end, so one pointing into the page header/offset table, or back into an already-parsed record, would read those bytes as a fabricated cookie. Offsets must now land past the offset table and records may not overlap any accepted so far. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The listing found Safari's jar with `stat`, which TCC permits without Full Disk Access, so Safari always listed as ready and the wizard opened on the configure step — the permission-first flow, the "still required" recheck and its `resume: "configure"` path could never run, and a refresh could not tell granted from denied. `unavailableReason` now probes the jar by opening it for read; TCC's EPERM reports `needsFullDiskAccess`, while a readable jar or any other failure does not. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The Safari reader set `domain` on every cookie, but `ImportedCookie.domain` is only for domain cookies (leading dot): Electron treats any supplied `domain` as one and re-adds the dot, widening host-only cookies to every subdomain and rejecting `__Host-` cookies outright. Route through the shared `cookieScope` like the Chromium and Firefox readers do, which also owns the IPv6 bracketing. Co-Authored-By: Claude Code <noreply@anthropic.com>

… denied macOS applies a Full Disk Access grant to the process identity at launch, so the running app can keep failing the probe after the toggle. Say so in the still-required panel. Fold the two recheck handlers into one, drop the unreachable no-localApi branch and its second toast, and document Safari import in the user guide. Co-Authored-By: Claude Code <noreply@anthropic.com>

github-actions · 2026-09-04T23:11:35Z

macOS preview

The preview download was removed because this PR closed or the preview label was removed.

- Add Safari profile discovery from SafariTabs metadata - Fall back to WebKit cookie stores when metadata is unavailable

cursor

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{Bugbot Autofix is ON, but it could not run because the branch was deleted or merged before autofix could start.}

^{Reviewed by Cursor Bugbot for commit c6cb164. Configure here.}

cursor · 2026-09-05T00:24:18Z

+    }
+  }
+  return profiles;
+});


FDA check skips named Safari jars

Medium Severity

listSafariProfiles can mark Safari installed from a named-profile jar, but the listing-time Full Disk Access probe still only opens the default "." jar. If that default jar is missing, TCC denial is not reported up front, so Safari looks ready or notInstalled instead of opening the permission step.

^{Reviewed by Cursor Bugbot for commit c6cb164. Configure here.}

github-actions Bot added vouch:trusted PR author is trusted by repo permissions or the VOUCHED list. size:L 100-499 changed lines (additions + deletions). labels Aug 16, 2026

macroscopeapp Bot reviewed Aug 16, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/SafariCookies.ts

juliusmarminge force-pushed the browser-import-safari branch from 88cf9f7 to 31db6ab Compare August 16, 2026 22:08

github-actions Bot added the 📱 Native Change Changes the native fingerprint; merging blocks production OTAs until a new store build ships. label Aug 16, 2026

macroscopeapp Bot reviewed Aug 16, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/SafariCookies.ts Outdated

juliusmarminge marked this pull request as ready for review August 16, 2026 22:18

cursor Bot reviewed Aug 16, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/SafariCookies.ts

Comment thread apps/desktop/src/preview/BrowserImport/SafariCookies.ts

juliusmarminge force-pushed the browser-import-safari branch from 31db6ab to 3df1bbe Compare August 16, 2026 22:53

juliusmarminge force-pushed the browser-import-safari branch from 3df1bbe to 79df949 Compare August 16, 2026 22:58

juliusmarminge force-pushed the browser-import-safari branch from 79df949 to d66d4fb Compare August 16, 2026 23:11

macroscopeapp Bot reviewed Aug 16, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/SafariCookies.ts

juliusmarminge force-pushed the browser-import-safari branch from d66d4fb to 06d5c22 Compare August 16, 2026 23:14

macroscopeapp Bot reviewed Aug 16, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/SafariCookies.ts

juliusmarminge force-pushed the browser-import-safari branch from 06d5c22 to 06722a1 Compare August 16, 2026 23:22

juliusmarminge force-pushed the browser-import-safari branch from 06722a1 to a8129c1 Compare August 16, 2026 23:25

cursor Bot reviewed Aug 16, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/Sources.ts

juliusmarminge force-pushed the browser-import-safari branch from a8129c1 to be13d9f Compare August 16, 2026 23:51

macroscopeapp Bot reviewed Sep 2, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/SafariCookies.ts

cursor Bot reviewed Sep 2, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/BrowserImport.ts Outdated

macroscopeapp Bot reviewed Sep 3, 2026

View reviewed changes

Comment thread apps/desktop/src/preview/BrowserImport/SafariCookies.ts

juliusmarminge mentioned this pull request Sep 3, 2026

fix(desktop): address the browser import review left over from the stack #9516

Merged

juliusmarminge and others added 20 commits September 4, 2026 14:19

fix(desktop): validate Safari cookie record headers

cda6152

fix(desktop): retain Safari cookie path on parse errors

289c96b

fix(desktop): open the Full Disk Access settings pane

0dcadc3

fix(web): recover initial Safari access

dc3c406

test(web): prioritize Safari access recovery

94f66f4

fix(web): report Safari access rechecks

46f6a53

fix(web): reject unavailable System Settings links

3833440

fix(web): explain denied Safari import retries

0945140

feat(desktop): discover Safari profile cookie stores

c6cb164

- Add Safari profile discovery from SafariTabs metadata - Fall back to WebKit cookie stores when metadata is unavailable

cursor Bot reviewed Sep 5, 2026

View reviewed changes

patroza mentioned this pull request Sep 5, 2026

merge: sync upstream through d2c3e2e5d7 (classic-merge) patroza/t3code#451

Merged

berghtho mentioned this pull request Sep 5, 2026

chore(upstream): sync 2026-09-05 berghtho/t3code#30

Merged

Uh oh!

Conversation

juliusmarminge commented Aug 16, 2026 • edited by macroscopeapp Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Add Safari cookie import with Full Disk Access permission flow

Uh oh!

coderabbitai Bot commented Aug 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review skipped

Uh oh!

Uh oh!

github-actions Bot commented Aug 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Thread transfer impact

Uh oh!

Uh oh!

Uh oh!

Uh oh!

macroscopeapp Bot commented Aug 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Approvability

Uh oh!

macroscopeapp Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

macroscopeapp Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

cursor Bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

macroscopeapp Bot commented Sep 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

github-actions Bot commented Sep 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

macOS preview

Uh oh!

cursor Bot left a comment

Choose a reason for hiding this comment

Uh oh!

cursor Bot Sep 5, 2026

Choose a reason for hiding this comment

FDA check skips named Safari jars

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

juliusmarminge commented Aug 16, 2026 •

edited by macroscopeapp Bot

Loading

coderabbitai Bot commented Aug 16, 2026 •

edited

Loading

github-actions Bot commented Aug 16, 2026 •

edited

Loading

macroscopeapp Bot commented Aug 16, 2026 •

edited

Loading

cursor Bot left a comment •

edited

Loading

macroscopeapp Bot commented Sep 3, 2026 •

edited

Loading

github-actions Bot commented Sep 4, 2026 •

edited

Loading